how does adsense determine pi? [newbie question]

After Search engine exclusion

[Ankhenaton]

hi i wondered how adsense determines pi impressions.

i get like 600 in the adsense form and from webalizer 3337 [right now slightly below half day]
this is after i threw out all bots page impressions. :/

yesterdays total in webalizer is 8386 but adsense shows 2.660.

When I go through my site, adsense is showing on all pages. If there is a malfunction I would expect to get statistically the same percentage on errors, PSA ads, empty fields etc :/.

Do so many people use ad blockers or am I doing something wrong? The main content is only up 4 days but the discrepancy remains. I don't have double pages etc.

If everybody else doesn't have the same experience I must do something wrong.

[wyweb]

I've never even tried to resolve server logs with adsense stats. I mean I saw right away there was discrepency and just moved on to something else.

It'll drive you crazy otherwise..

[Ankhenaton]

hmm, i will probably be forced to ignore it, but a 1:4 error is quite significant and I thought worth investigating.

[nesalc]

Internet explorer's caching could explain it.

This is the reason why you can have 2 or more clicks and only one page impression.

/Claus

[Ankhenaton]

Internet explorer's caching could explain it.
This is the reason why you can have 2 or more clicks and only one page impression.

I use a php mysql software with page caching, that's possibly it together with client side caching.
Since I assume the clicks are real, i was wondering why the ctr is so high. When i correct by 4:1 it's more believable, I guess.

i just clicked a couple of times simple reload and the ads are changing .. beats me ..

[nesalc]

I have never been able to figure out the exact behavior of internet explorer.

Sometimes when you click back, contents is reloaded, like when pressing f5, but other times it just loads stuff from the cache. And I'm talking behavior for the same page.

/Claus

[arrowman]

i wondered how adsense determines pi

Google: pi * 1
Or: pi + 0

[xtreem]

or better, "pi =" or "what is pi"

[queritor]

Google only shows 'real' page views initiated by a real browser with javascript enabled.

Are you sure you are comparing 'like for like'?

Although Webalizer will correctly exclude images, it shows every single page request, every form submission, every css file, every 'included' file and anything else that's not an image. It doesn't care whether the file request results in a 'real' page view.

I use Urchin which uses a javascript tag to track page views. My stats match all the third party ad networks within 10%.

[Ankhenaton]

Are you sure you are comparing 'like for like'?

Although Webalizer will correctly exclude images, it shows every single page request, every form submission, every css file, every 'included' file and anything else that's not an image. It doesn't care whether the file request results in a 'real' page view.

You can decide what is a pageview in webalizer as well as ignore certain domains. I therefore excluded googlebot, inktomi etc, myself, my business partner etc. I also excluded gif, jpeg, jpg, png .js css etc.

Anyway I am looking for a better analysis tool webalizer is a bit old now. :)